Doh! You are right. The others are probably not looking at it because it doesn't
make any sense. I did for mean versions 2 and 4 to have s1+s2, and not s2+s2...
and then it all works fine. That's what you get for working late into a Saturday
evening (in the UK) instead of having a life!
Cheers!
Paul
-----Original Message-----
From: Walt Brainerd <[log in to unmask]>
To: [log in to unmask]
Date: Sat, 17 May 2003 22:20:30 -0700
Subject: Re: WHERE and MATMUL combination
I'll take a shot at this, since others seem to have
a life and aren't looking at this.
I am not completely sure I understand your concern.
But one thing odd: did you mean versions 2 and 4 to
have s2+s2 instead of s1+s2? If you make that change
you should get the same result in all four cases.
I am not sure if this is related to your problem,
but MATMUL inside a WHERE is completely evaluated
*before* the mask is applied (the mask does not
apply to anything "inside" a transformational
function.) Thus you get the same result as in the
other cases (with the above "fix").
Lemme know if I completely misunderstand your concern.
|